SU‐E‐T‐622: Planning Technique for Passively‐Scattered Involved‐Node Proton Therapy of Mediastinal Lymphoma with Consideration of Cardiac Motion

Abstract

Purpose:To develop a treatment planning method for passively‐scattered involved‐node proton therapy of mediastinal lymphoma robust to breathing and cardiac motions.Methods:Beam‐specific planning treatment volumes (bsPTV) are calculated for each proton field to incorporate pertinent uncertainties. Geometric margins are added laterally to each beam while margins for range uncertainty due to setup errors, breathing, and calibration curve uncertainties are added along each beam. The calculation of breathing motion and deformation effects on proton range includes all 4DCT phases. The anisotropic water equivalent margins are translated to distances on average 4DCT. Treatment plans are designed so each beam adequately covers the corresponding bsPTV. For targets close to the heart, cardiac motion effects on dosemaps are estimated by using a library of anonymous ECG‐gated cardiac CTs (cCT). The cCT, originally contrast‐enhanced, are partially overridden to allow meaningful proton dose calculations. Targets similar to the treatment targets are drawn on one or more cCT sets matching the anatomy of the patient. Plans based on the average cCT are calculated on individual phases, then deformed to the average and accumulated. When clinically significant dose discrepancies occur between planned and accumulated doses, the patient plan is modified to reduce the cardiac motion effects.Results:We found that bsPTVs as planning targets create dose distributions similar to the conventional proton planning distributions, while they are a valuable tool for visualization of the uncertainties. For large targets with variability in motion and depth, integral dose was reduced because of the anisotropic margins. In most cases, heart motion has a clinically insignificant effect on target coverage.Conclusion:A treatment planning method was developed and used for proton therapy of mediastinal lymphoma. The technique incorporates bsPTVs compensating for all common sources of uncertainties and estimation of the effects of cardiac motion not commonly performed.